A distributed computing or processing system comprises multiple computers, such as compute nodes or processing nodes, which operate relatively independently to achieve a common goal. Data processing, analytics, and/or other jobs may be performed using distributing computing (e.g., hundreds or thousands of compute nodes may process a relatively large data set to accomplish a job). A scheduler may be used to schedule tasks, corresponding to portion of a job, which are to be performed by various compute nodes.